In three-stage prospective cohort controlled research of 2876 persons with chronic diffuse liver diseases (2076 cases of fatty liver disease, 509 cases of chronic hepatitis, 139 cases of liver cirrhosis) and 152 healthy controls the clinical significance of radiation methods of investigation in quantitative evaluation of liver morphofunctional state in patients with chronic diffuse liver diseases was estimated. Diagnosisprognostic algorithm for patients with this pathology was developed, validated and recommended for practical application. Further perspectives are related to detailed chronic diffuse liver diseases stratification by intensity of steatosis (in fatty liver disease), activity (in chronic hepatitis) and fibrosis (in chronic hepatitis and liver cirrhosis).
ROLE OF ECHOGRAPHY AND COMPUTED TOMOGRAPHY IN DIAGNOSIS OF CHRONIC DIFFUSE LIVER DISEASES
1. Pharmaceutical and Biomedical Sciences: An International Journal (PBIJ), Vol.1, No.1
9
ROLE OF ECHOGRAPHY AND COMPUTED
TOMOGRAPHY IN DIAGNOSIS OF CHRONIC
DIFFUSE LIVER DISEASES
Yuliya Ya. Fedulenkova
Department of Radiology and Radiation Medicine, Kharkiv National Medical
University, Kharkiv, Ukraine
ABSTRACT
In three-stage prospective cohort controlled research of 2876 persons with chronic diffuse liver diseases
(2076 cases of fatty liver disease, 509 cases of chronic hepatitis, 139 cases of liver cirrhosis) and 152
healthy controls the clinical significance of radiation methods of investigation in quantitative evaluation of
liver morphofunctional state in patients with chronic diffuse liver diseases was estimated. Diagnosis-
prognostic algorithm for patients with this pathology was developed, validated and recommended for
practical application. Further perspectives are related to detailed chronic diffuse liver diseases
stratification by intensity of steatosis (in fatty liver disease), activity (in chronic hepatitis) and fibrosis (in
chronic hepatitis and liver cirrhosis).
KEYWORDS
Fatty Liver Disease, Chronic Hepatitis, Liver Cirrhosis, Ultrasound Diagnosis, Computed Tomography
1. INTRODUCTION
Chronic diffuse liver diseases cause not only significant medical problem, promoting portal
hypertension, hepatocellular failure, oncological diseases etc. [1], but also social and economical,
associated with loss of ability to work in active population; particularly, mortality due to liver
cirrhosis in European countries vary from 100 to 400 per 1 million of male population and from
40 to 150 per 1 million among females [2, 3, 4, 5].
In last decade is an impetuous development of chronic diffuse liver disease takes place.
Technical progress, improvement of hardware and software have made it possible to significantly
go forward in understanding of etiology and pathogenesis of hepatobiliary pathology, noticeably
improve treatment results. In many respects these changes became possible only due to
introduction and development of new highly-technological methods of radiation (visual)
diagnosis in general and ultrasonography – as well [6, 7].
Results of radiological diagnostic methods are widely used in order to build prognostic models
and calculate a prognosis in each certain clinical case about functional compensation of liver,
presence of indications for radical intervention etc. [8]. Both mathematical [9, 10, 11, 12] and
virtual [13], dimensional models based on morphological investigation of liver come in handy
and give an opportunity to more exactly and objectively evaluate changes which develop in vivo,
formulate an individual prognosis for a patient.
2. Pharmaceutical and Biomedical Sciences: An International Journal (PBIJ), Vol.1, No.1
10
Despite the availability of wide spectrum of other modern non-invasive and low-invasive
radiation methods of diagnosis for chronic diffuse liver diseases, ultrasonography does not lose
its topicality and is the leader of practical application in public health service. Undoubted
perspective is characteristic for methods which base not only on immediate primary results of
investigation, but on results of their further objectivization using mathematical modeling and
building of prognostic model of the pathological process. Reasonability of ultrasonography
application for diagnosis of chronic diffuse liver diseases is related to its following
characteristics: non-invasive [14], safe [15], easily done [16], relatively cheap and capable to
simultaneous evaluation of comorbid pathology [17], also in dynamics, reproductible and
precise, standardized [18], easily combined with other methods [19, 20], capable for further
improvement and including contrast-enhance.
While, some limitation of ultrasonography in diagnosis of chronic diffuse liver diseases takes
place in medical practice [21, 22]. This problem may be solved by improvement of
ultrasonography diagnosis approach with automation of image analysis [23], introduction of
index parameters etc.
Thus, despite general success of ultrasonography in diagnosis of chronic diffuse liver diseases,
till nowadays improving of this approach remains vital.
2. MATERIAL AND METHODS
Aim of research – optimization of diagnosis of chronic diffuse liver disease by development of
diagnosis-prognostic algorithm using USI and CT parameters.
On first phase an estimation of diagnostic capabilities (values) of ultrasonography in chronic
diffuse liver diseases was performed. It included investigation of 103 cases of chronic diffuse
liver disease: fatty liver disease (n=63), chronic hepatitis (n=12), liver cirrhosis (n=28), cases
with intact liver (control group, n=2). This phase was conducted on the base of Kharkiv Regional
Clinical Hospital – Center of Emergency Medical Care and Disaster Medicine in 2012–2013
with the aim of evaluation of accordance of ultrasonographic data with results of liver autopsy
(both n=105).
Second phase was aimed on development of a diagnosis-prognostic algorithm of ultrasonography
application and profound instrumental-laboratory tests (apart of ultrasound – anamnestic and
physical data collection, complete blood count and biochemical blood tests) were performed.
Two hundred fifty three persons, including patients with fatty liver disease (n=142), chronic
hepatitis (n=43), liver cirrhosis (n=27), and 50 healthy people shared this phase of the research. It
has been conducted in clinical hospital of Grigoriev Institute of Medical Radiology of National
Academy of Medical Science of Ukraine in 2013–2014. All participants underwent detailed
repetitious instrumental and laboratory investigations. In this cohort male and female age
medians did not significantly differ.
Third phase was most large in the research as aimed on evaluation of ultrasonography
capabilities in prospective evaluation of chronic diffuse liver disease course by validization of
developed diagnosis-prognostic algorithm for pathologic process severity evaluation and course
prediction, and also building a mathematic model of pathologic process. In order to reach the
goal, 2206 persons were examined, including patients with fatty liver disease (n=1660), chronic
hepatitis (n=416), liver cirrhosis (n=45), control healthy individuals (n=85). This phase of the
research was performed on the base of LTD Medical Diagnostic Center «Expert-Kharkov» in
2013–2014. Age median was 56.1 (47.1; 64.1) years, 55.0 (44.3; 64.2) years in men, 56.4 (47.7;
64.1) years in women (difference is reliable by р=0,053).
3. Pharmaceutical and Biomedical Sciences: An International Journal (PBIJ), Vol.1, No.1
11
Ultrasound investigation was performed with Xario SSA 660A system (Toshiba Medical
Systems, Japan). Pathomorphological investigation has been conducted according to standard
protocols with evaluation, where applicable, of liver steatosis [24] and liver fibrosis intensity
degree [25], level of activity (Knodell R. G., 1981). Nonparametric methods of statistical
analysis were applied [26]. Median (Mе) and interquartile interval with representing lower,
25 %, quartile (LQ) and upper, 75 % quartile (UQ) were calculated, result was expressed by
Me (LQ; UQ) way for shortness. Kruskal-Wallis ANOVA & median test method, Mann-
Whitney U-criterion, Wilcoxon method, correlation by Spearman, Fisher angular transformation
were used where applicable. Comparative analysis in groups of separate diagnostic criteria
distribution using ANOVA and Wald sequential analysis (Wald A., 1947) in its interpretation for
medical diagnosis (Genkin A. A., 1962; Gubler E. V., 1978) by ranging of parameters by their
differential-diagnostic information capacity allowed to define diagnostic value, prognostic
significance and influence power of factors on parameters divergence in clinical groups and
prognostic coefficients. Only independent prognostic parameters were included in the algorithm.
In cases when correlation strengths between factors were more than |0,70|, one of factors was
excluded from list of parapeters. Ranking of parameters by influence strengths, evaluation of
prognostic and information capacity of parameters allowed choosing the most reliable
parameters. At the last phase, mathematic modeling using discriminant analysis and building of
artificial virtual neural networks with their further training was performed.
The research has been conducted according to requirements of European Convention (Strasburg,
1986), ІСН GСP (2008), GLР (2002) principles, local Ukrainian legislation.
3. RESULTS AND DISCUSSION
In 1st phase the age median was 62 (46; 71) years without significant gender differences, while
gender comparisons in different nosologic groups showed the age difference (р=0.0001) – the
oldest in fatty liver disease group, 64.0 (60.0; 73.0) years; youngest in chronic hepatitis group,
41.6 (31.0; 52.5) years; intermediate in liver cirrhosis group, 52.0 (41.5; 65.5) years. Age
comparison in different nosologic groups in further phases showed above mentioned (in 1st
phase) differences in general and among women (both р<0.001).
Pathomorphological verification of ultrasonography results in evaluation of disease character in
patients with chronic diffuse liver diseases revealed 1.9 % probability of false-positive
ultrasound diagnosis of this pathology, diagnostic capacity of 14–97 % (with central value of
56 %).
Using results of profound clinical, instrumental and laboratory investigation including frequency
of separate ultrasonography parameters and prognostic value of each of criterion the screening
algorithm was elaborated in order to predicting of complication risk. It has table form, which
includes demographic-antropometric (age, gender, body mass index), sonographic (liver size,
characteristics of capsule, parenchyma, ascites, hepatic vein circulation, caudal to right lobe
transverse size, degree of steatosis, congestion index, modified hepatic index, hepatic vascular
index, index of arterial perfusion, portolienal venous index, pulsatory index of spleen artery,
platelets to spleen diameter ratio, right lobe width to albumins ratio) parameters, appropriate
prognostic coefficients and prognostic result evaluation scale. By each parameter its presence or
absence has to be evaluated, corresponding prognostic coefficients are to be summarized. By
achievement of threshold sum of coefficients a risk group was stated by using the scale: if equal
or less than -19.8, risk is minimal; if more than -19,8 and less than 19.8, risk is uncertain; if equal
or more than 19.8, risk is high.
4. Pharmaceutical and Biomedical Sciences: An International Journal (PBIJ), Vol.1, No.1
12
For each of three diseases canonic discriminant functions were built. Predicted appliance to
groups of low, uncertain, high progression risk in fatty liver disease was 78.9 %, 69.3 %, 99.7 %
accordingly; in chronic hepatitis – 63.7 %, 61.1 %, 93.2 % accordingly; in liver cirrhosis –
93.1 %, 64.8 %, 99.7 %.
Neural networks (three-level perceptron with descending number of nodes) were built, trained
and sensitivity, specificity evaluated. After training the square below ROC-curve to each of risk
groups became increased > 80 %.
4. CONCLUSIONS
1. Prospective randomized three-phase populational (on 3rd phase) research has proved the
clinical significance of ultrasonography in evaluation of morphofunctional state of liver and
portohepatolienal bloodstream in chronic diffuse liver diseases (fatty liver, chronic hepatitis,
liver cirrhosis).
2. Diagnosis-prognostic algorithm which includes anthropodemographic, clinical and
ultrasonographic parameters has been elaborated, intending on forecast of risk of unfavourable
course of chronic diffuse liver diseases.
3. Predicting value of ultrasonography has been boosted by discriminant mathematic model
development, artificial virtual neural networks application.
4. Longitude multicenter study of proposed approach in diagnosis of chronic diffuse liver disease
using non-invasive ultrasonography and evaluation of treatment efficacy might be further
perspectives of the research.
ACKNOWLEDGEMENTS
The author is thankful to administration of Kharkiv National Medical University; Kharkiv
Regional Clinical Hospital – Center of Emergency Medical Care and Disaster Medicine;
Grigoriev Institute of Medical Radiology of National Academy of Medical Science of Ukraine,
Medical Diagnostic Center «Expert-Kharkov», LTD for support in the research for this initiative
medical research of the Department of Radiology and Radiation Medicine as well as individual
Ph. D. dissertation research of the author.
REFERENCES
[1] Mauss S, Berg T, Rockstroh J, Sarrazin C & Wedemeyer H, editors (2015), Hepatology 2015,
Sydney, Flying Publisher
[2] Byrne CD & Targher G, (2015) “NAFLD: A multisystem disease”, J Hepatol, Vol. 62, No. 1,
Suppl., ppS47-S64.
[3] Mathurin P & Bataller R, (2015) “Trends in the management and burden of alcoholic liver
disease”, J Hepatol, Vol. 62, No. 1, Suppl., ppS38-S46.
[4] Stickel F, (2015) “Alcoholic cirrhosis and hepatocellular carcinoma”, Adv Exp Med Biol, Vol.
815, pp113-130.
[5] Mokdad AA, Lopez AD, Shahraz S, Lozano R, Mokdad AH, Stanaway J, Murray CJ & Naghavi
M, (2014) “Liver cirrhosis mortality in 187 countries between 1980 and 2010: a systematic
analysis”, BMC Med, Vol. 12, No. 1, pp145.
5. Pharmaceutical and Biomedical Sciences: An International Journal (PBIJ), Vol.1, No.1
13
[6] Aubé C, (2015) “Imaging modalities for the diagnosis of hepatic fibrosis and cirrhosis”, Clin Res
Hepatol Gastroenterol, Vol. 39, No. 1, pp38-44.
[7] Gerstenmaier JF, Gibson RN, (2014) “Ultrasound in chronic liver disease”, Clin Res Hepatol
Gastroenterol, Vol. 39, No. 1, pp38-44.
[8] Hassan EA, Abd El-Rehim AS, (2013) “A revised scope in different prognostic models in
cirrhotic patients: Current and future perspectives, an Egyptian experience”, Arab J
Gastroenterol, Vol. 14, No. 4, pp158-164.
[9] Huhdanpaa H, Douville C, Baum K, Krishnamurthy VN, Holcombe S, Enchakalody B, Wang L,
Wang SC & Su GL, (2011) “Development of a quantitative method for the diagnosis of
cirrhosis”, Scand J Gastroenterol, Vol. 46, No. 12, pp1468-1477.
[10] Krishnamurthy V, Zhang P, Ethiraj S, Enchakalody B, Waljee AK, Wang L, Wang SC & Su GL,
(2015) “Use of analytic morphomics of liver, spleen, and body composition to identify patients at
risk for cirrhosis”, Clin Gastroenterol Hepatol, Vol. 13, No. 2, pp360-368.
[11] Gao S, Peng Y, Guo H, Liu W, Gao T, Xu Y & Tang X, (2014) “Texture analysis and
classification of ultrasound liver images”, Biomed Mater Eng, Vol. 24, No. 1, pp1209-1216.
[12] Pournik O, Dorri S, Zabolinezhad H, Alavian SM & Eslami S, (2014) “A diagnostic model for
cirrhosis in patients with non-alcoholic fatty liver disease: an artificial neural network approach”,
Med J Islam Repub Iran, No. 28, pp116.
[13] Chen YW, Luo J, Dong C, Han X, Tateyama T, Furukawa A & Kanasaki S, (2014) “Computer-
aided diagnosis and quantification of cirrhotic livers based on morphological analysis and
machine learning”, Comput Math Methods Med, doi: 10.1155/2013/264809.
[14] Heller MT & Tublin ME, (2014) “The role of ultrasonography in the evaluation of diffuse liver
disease”, Radiol Clin North Am, Vol. 52, No. 6, pp1163-1175.
[15] Reiser MF, Hricak H & Knauth M, authors; Maconi G & Porro GB, editors (2014), Ultrasound of
the gastrointestinal tract, Berlin Heidelberg, Springer
[16] Dhamija E, Thulkar S & Bhatnagar S, (2015) “Utility and potential of bedside ultrasound in
palliative care”, Indian J Palliat Care, Vol. 21, No. 2, pp132-136.
[17] Jepsen P, (2014) “Comorbidity in cirrhosis”, World J Gastroenterol, Vol. 20, No. 23, pp7223-
7230.
[18] Neciu C, Badea R, Chiorean L, Badea AF & Opincariu I, (2015) “Oral and I.V. Contrast
Enhanced Ultrasonography of the digestive tract – a useful completion of the B-mode
examination: a literature review and an exhaustive illustration through images”, Med Ultrason,
Vol. 17, No. 1, pp62-73.
[19] Cobbold JF, Crossey MM, Colman P, Goldin RD, Murphy PS, Patel N, Fitzpatrick J, Vennart W,
Thomas HC, Cox IJ & Taylor-Robinson SD, (2010) “Optimal combinations of ultrasound-based
and serum markers of disease severity in patients with chronic hepatitis C”, J Viral Hepat, Vol.
17, No. 8, pp537-545.
[20] Fan M, Su C, Lu L, Qin JC, Li P & Yuan JY, (2014) “Ultrasonic diagnosis and vasoactive
substances examination in patients with cirrhosis”, Asian Pac J Trop Med, Vol. 7, No. 4, pp329-
332.
[21] Soresi M, Giannitrapani L, Cervello M, Licata A & Montalto G, (2014) “Non invasive tools for
the diagnosis of liver cirrhosis”, World J Gastroenterol, Vol. 20, No. 48, pp18131-18150.
[22] Rumack C, Wilson S, (2011), Diagnostic Ultrasound, Philadelphia, Mosby
[23] Vicas C, Lupsor M, Socaciu M, Nedevschi S & Badea R, (2012) “Influence of expert-dependent
variability over the performance of noninvasive fibrosis assessment in patients with chronic
hepatitis C by means of texture analysis”, Comput. Math. Methods. Med, doi:
10.1155/2012/346713.
6. Pharmaceutical and Biomedical Sciences: An International Journal (PBIJ), Vol.1, No.1
14
[24] Chirsky VS, author; Khmelnitskaya NM, editor (2009), Biopsy diagnosis of non-tumorous liver
diseases, Saint-Petersburg
[25] Roux S, Tournayre J, Mahul A, Debroas D, Enault F, (2014) “Metavir 2: new tools for viral
metagenome comparison and assembled virome analysis”, BMC Bioinformatics, No. 15, pp76.
[26] Lang TA, Secic M, (2006) How to report statistics in medicine: Annotated guidelines for authors,
editors, and reviewers, Philadelphia, American College of Physicians.
Authors
Fedulenkova Yuliya Yanovna, radiologist, researcher & teacher of the
Department of Radiology and Radiation Medicine of Kharkiv National
Medical University